INDEX

Explanations

references to sources of information and authority figures in a discussion or analysis

oai_token-act-pair · gpt-4o-mini Triggered by @bot

entities reporting or estimating

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

"""

-0.39

“

-0.37

 Hang

-0.37

 where

-0.36

lla

-0.36

myapplication

-0.35

 burning

-0.35

Nas

-0.34

Здравствуйте

-0.34

どこか

-0.33

POSITIVE LOGITS

 Administrativna

0.66

AndEndTag

0.56

verwijspagina

0.54

cèse

0.51

 theorists

0.50

 esperti

0.50

 cejas

0.50

 otomatig

0.50

 varandra

0.49

 مرئيه

0.49

Activations Density 0.050%

references to sources of information and authority figures in a discussion or analysis

entities reporting or estimating

No Comments

No Known Activations